Re: If anybody wants to make something in Ada but do not know what

["Warren W. Gay VE3WWG" <ve3wwg@cogeco.ca>]

Preben Randhol wrote:
> Then perhaps a Bayesian Spam filter could be a nice challenge. Or if
> somebody are heading a university student project/diploma work it could
> a suitable project? 
> 
> Spam is becoming a great problem. Last check I got 30 spam messages in 3
> days that was caught by my filter (spamassassin) and about 10 that
> wasn't. There was also about 3 false negatives (mailed labelled as SPAM
> but which was really HAM). So I think it is time for me to change spam
> filtering method. Besides spamassassin is a perl script program and it
> is a hug resource hog on my computer and it is slow. There are two
> Bayesian Spam filter implementations in C (ugg :-( ) which has the
> advantage of being fast. Problem is that with C's track record on buffer
> overflow I don't trust them to be secure. This is why an Ada
> implementation would be nice I think. Besides spam filters is something
> people need so it could promote Ada95.
> 
> Wish I could make it myself, but I simply don't have the time at the
> moment.
...
> Preben

I personally believe that as long as email is free (and we like it
that way), SPAM will continue to be a big problem. To keep email
free, I think that the only way this will work is we'll end up
using two forms of email on the Internet:

   1) Traditional email, which permits unsolicited mail (port 25);
      which will continually battle with spam filters and such.
   2) A newer "solicited only" email system (port 26?)

Perhaps the 2nd kind, can piggy back off of the first, by either
using a similar but extended protocol, and probably
using a new TCP/IP port # to avoid the log jam that occurs
on port 25 as spammers spam away.

The new email protocol would make use of PKI exchanges (PGP?),
so that only those people that you have shared certificates
with, can successfully deposit email in your mailbox (this
would be great for kids, so that only their friends can
send them mail etc.). If someone gives away the cert (if the
protocol allows it), then you revoke it, and issue a new one
if necessary.

Where the protocol research comes in (I think), is the method of
sharing and administering certs in a way that is easy for
grandmothers to work with.

However, this project probably needs to start with an IETF
committee, but certainly some Ada prototypes can be useful
as "proof of concept" projects.

Like you Preben, I'd do this myself if I had the time, but
I have to face my own limitations and mortality ;-)

Even getting the IETF interested in a formal proposal requires
too much of my time to launch. Maybe someone else is willing
to think it through and run with it. They can even claim it
as their own idea ;-)

-- 
Warren W. Gay VE3WWG
http://home.cogeco.ca/~ve3wwg